Genomic Studies Investigate Treatment Response and Inherited Risk in Sarcoma

Tempus AI (Chicago, IL, USA) will present six abstracts, including three oral presentations, at the Connective Tissue Oncology Society (CTOS) 2026 Annual Meeting, taking place November 18–21 in Seoul, South Korea. The studies examine molecular biomarkers, tumor classification, treatment response, and genetic alterations in rare cancers, including sarcomas and gastrointestinal stromal tumors (GIST).

One study examines melanoma-associated antigen A4 (MAGE-A4) messenger RNA (mRNA) expression and the tumor immune microenvironment as potential biomarkers in sarcoma. Using real-world clinical and genomic data, researchers are investigating how gene expression and immune characteristics may help identify biologically distinct tumors and inform treatment selection.

Image Credit: Adobe Stock

Two additional studies focus on GIST. One identifies a DNA damage response (DDR) transcriptional signature enriched in a subset of succinate dehydrogenase (SDH)-deficient tumors, with potential implications for sensitivity to temozolomide. Another uses transcriptomic profiling of 1,490 patients to classify GIST into prognostic risk groups, examining whether gene expression patterns can improve risk assessment beyond conventional clinical and molecular variables.

Another presentation examines transcriptional programs associated with response to pazopanib in patients with metastatic bone and soft tissue sarcoma. The research investigates gene expression patterns linked to treatment outcomes, with the aim of identifying molecular characteristics that may help predict which patients are more likely to benefit from targeted therapy.

Additional research examines the molecular and genomic characteristics of desmoplastic small round cell tumors, a rare and aggressive sarcoma subtype. A separate real-world analysis involving 9,025 sarcoma patients evaluates matched tumor-normal sequencing to improve the identification of inherited genetic variants that may otherwise be difficult to distinguish from tumor-specific alterations.

The studies draw on Tempus' de-identified multimodal dataset, which includes 45,085 patients across 615 rare and ultra-rare cancer types. Sarcomas are among the most extensively represented cancers in the dataset. The research combines matched tumor-normal DNA sequencing, whole-transcriptome analysis, and digital pathology information to investigate molecular characteristics associated with disease behavior and treatment response.

Tempus uses molecular profiling, artificial intelligence (AI), and clinical data analysis to support precision oncology research and diagnostic interpretation. Its Orphan Disease Program focuses on challenges associated with rare cancers, where limited patient populations can make biomarker discovery and clinical research difficult. The broader approach aims to identify molecular patterns that may inform diagnosis and therapeutic development.

Related Links
Tempus AI