Abstract 2849: Comparative analysis of mutational signatures in TumorGraft PDX bank and TCGA dataset to improve patient tumor modeling

Abstract Mutational signatures are unique patterns of DNA alterations that arise from specific mutagenic processes, such as exposure to genotoxic agents or DNA repair deficiencies. Tumors with common mutational signatures can harbor defects in specific DNA repair pathways that may underlie common genetic vulnerabilities. Furthermore, exposure patterns of mutational signatures can vary among clinical types of cancer. Champions’ TumorGraft platform, a collection of 1500 patient-derived xenografts generated from more than 50 different types of cancer, is one of the most comprehensive preclinical oncology platforms worldwide and aims to recapitulate the variety of the patient population and tumor biology complexity. We performed the mutational signatures by profiling our models through whole exome sequencing. The results obtained were crossed with the patient’s clinical history and showed a good correlation between the mutational signatures identified and tumor etiology and DNA damage repair alterations known to be associated with some indications. To assess the clinical relevance of our findings, we applied the molecular signature profiling method used with our PDXs to the TCGA dataset. This analysis revealed that the mutational profiles enabled the co-clustering of patients' tumors and PDXs from the same indications. As expected, we also observed co-clustering of tumors from indications that share exposure to a specific environmental or therapeutic genotoxic agent (e.g., tobacco smoke, chemotherapy, or radiation), or common pro-oncogenic genetic events associated with DNA damage repair alterations, such as homologous recombination repair, that can be detected in tumor samples across different indications. In addition to this, we found other signatures, associated with known, inferred or unknown etiology, detected across indications and affecting a limited number of tumors per indication. By combining the insights from these mutational signatures with the clinical data and the comprehensive pharmacological and molecular characterization of our models, including whole transcriptome, proteomics, and phospho-proteomics data, we evaluated the association of the tumors that harbor these signatures with tumor biology and clinical outcome. This approach led to the identification of mechanisms of action and biomarkers that might allow the identification of personalized therapies for patients with these tumors that are driven by similar mutational profiles, regardless of tumor indication. Citation Format: Stefano Cairo, Mara Gilardi, Marianna Zipeto, Michael Ritchie, Gilad Silberberg. Comparative analysis of mutational signatures in TumorGraft PDX bank and TCGA dataset to improve patient tumor modeling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2849.