Abstract 688: Characterization of microsatellite status in endometrial tumors using comprehensive genomic profiling via the PGDx elio tissue complete

Abstract Background: Microsatellite instability (MSI) is a vital genomic biomarker in endometrial cancer (EC), playing a key role in guiding patient management decisions. Microsatellite status can be assessed using various methods, including immunohistochemistry (IHC), polymerase chain reaction (PCR), and next-generation sequencing (NGS). However, accurately determining microsatellite status in EC can be challenging due to the typically weak MSI phenotype observed in this cancer type. This study aims to enhance microsatellite analysis in EC by combining EC mutation profiling with MSI detection techniques into a unified analytical workflow. Methods: We performed comprehensive genomic profiling on 43 EC samples using the PGDx elio™ tissue complete (ETC), an FDA-cleared NGS test targeting 505 solid tumor related genes. In addition to mutation profile, we obtained microsatellite status and tumor mutational burden for each case. Mismatch repair (MMR) status were also determined for all 43 ECs using IHC (MLH1, PMS2, MSH2 and MSH6). ‘maftools’ R package was utilized for data processing and visualization, while genomic classifier was developed using ‘randomForest’ R package. Results: IHC analysis led to identification of 28 proficient MMR (pMMR) and 15 deficient MMR (dMMR) EC samples. We observed 100% (28/28) concordance for pMMR samples and 73.3% (11/15) concordance for dMMR samples between IHC analysis and NGS assay. Four discordant samples showed low TMB and low-grade instability in 68 mononucleotide tracts considered by NGS assay indicative of weak MSI phenotype. We also identified 15 genes (including ERBB4, PTPRT, ARID1A, ARID1B, and BRD4) significantly altered in dMMR samples compared to pMMR samples. Random forest classifier was developed considering 88 distinct parameters (including mutation profile of 15 genes, 5 clinicopathologic features and 68-mononucleotide tract status) for all EC samples. Conclusion: We observed varying degrees of concordance (100% specificity and 73.3% sensitivity) between MSI status by NGS based ETC and MMR status by IHC in a cohort of EC. Targeted genomic profiling via ETC identified significantly altered genes in dMMR endometrial tumors and led to development of random forest classifier to predict MSI EC tumor [which showed AUC value of 0.875, 87.5% specificity and 100% sensitivity for test set]. Integration of mutational profiling with MSI detection can be leveraged to increase sensitivity for MSI detection on the ETC platform. Citation Format: Darshan Shimoga Chandrashekar, João Lobo, Mehenaz Hanbazazh, Kanako Okamoto, Jennifer B. Jackson, Shuko Harada, Alexander C. Mackinnon. Characterization of microsatellite status in endometrial tumors using comprehensive genomic profiling via the PGDx elio tissue complete. [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 688.