Abstract 2702: Integrative Computational Pipeline for Tracking Neoantigen Retention Across PDX Passages in Genitourinary Cancers.

Abstract Cancer neoantigens are highly attractive targets for personalized immunotherapies because they are uniquely expressed on tumor cells and absent from normal tissues. However, validating predicted neoantigens—through immunogenicity testing, HLA characterization, and expression analysis—is often limited by the small amount of tumor tissue obtainable in clinical settings. Patient-derived xenograft (PDX) models therefore serve as valuable systems for expanding tumor material and studying tumor evolution, yet the extent to which PDX models retain patient-specific neoantigens across passages remains insufficiently characterized. In this pilot study, we analyzed neoantigen prediction and retention in five (n=8) genitourinary tumors and their matched PDX passages using paired whole-exome sequencing (WES) and RNA-seq data from the NCI Patient-Derived Models Repository (PDMR). We implemented a reproducible, semi-automated Nextflow workflow combining nf-core pipelines for tumor-normal variant calling (sarek), HLA class I typing (hlatyping), transcript quantification (rnaseq), and in silico HLA-peptide binding prediction (epitopeprediction). Variant clonality was assessed with PureCN, and downstream analyses prioritized expressed, nonsynonymous variants with predicted HLA-binding affinity. Primary tumors contained 14-156 candidate neoantigens per patient. Retention in derived PDX models was generally high: initial engraftment (P0) preserved 50-89% of primary tumor neoantigens, and later passages maintained 46-95%. Although some neoantigens were lost during engraftment or propagation, most remained stable over serial passages, indicating preservation of key features of the tumor-specific immunogenic landscape. By the time of presentation, this analysis will be expanded beyond the initial five cases to include a larger PDMR cohort of genitourinary cancers. We will also incorporate our own PDX and CDX datasets to evaluate neoantigen retention or divergence across passages in an independent system. Together, these results will provide a broader assessment of neoantigen stability in patient-derived models and support the identification of robust, persistent neoantigens for personalized vaccine development. Citation Format: Md Imran Khan, Tyler Gross, Christian Migliarese, Ian Shea, Jonathan F. Lovell, Roberto Pili, . Integrative Computational Pipeline for Tracking Neoantigen Retention Across PDX Passages in Genitourinary Cancers abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2702.