Abstract 2486: Decoding tumor evolution through fitness landscapes of aneuploid cells

Abstract Aneuploidy, prevalent in most solid tumors, significantly alters cellular phenotype and fitness. Despite its critical role in tumor evolution, a detailed quantitative understanding of its evolutionary dynamics remains elusive. To address this gap, we developed ALFA-K, a novel method to infer the fitness landscape of chromosome-level karyotypes. Utilizing longitudinal single-cell karyotype data from evolving cell populations, our method estimates the fitness of thousands of karyotypes proximal to the input data in karyotype space, enabling predictions of unseen karyotypes. We validated our approach using both synthetic data from an agent-based model (ABM) and empirical data from in vitro and in vivo passaged cell lines. The empirical data comprised over 40, 000 sequenced cells across five cell lines, each sequenced longitudinally at 5-10 timepoints per lineage. Analysis of the fitted landscapes yielded several key insights: 1) WGD alters the distribution of fitness effects of copy number alterations (CNAs) (K.S. test: P<10-). Specifically, karyotypes with WGD exhibit more mutations with positive selection coefficients compared to diploid karyotypes; 2) Mutations in cisplatin-treated lineages showed changes in fitness effects on average 2.3 times larger than those in untreated lineages (t-test: P<2.2×10-1); 3) The fitness impact of a CNA varies depending on the parental karyotype; 4) The rate of chromosome missegregation can dictate the predominant karyotypes in an evolving population. To further investigate how the tumor microenvironment shapes karyotype evolution, we are integrating these inferred fitness landscapes into an existing spatial ABM. This modeling simulates the spatially heterogeneous tumor microenvironment and its influence on karyotypic selection and adaptation. By leveraging this ABM, we aim to capture the effects of spatial constraints—such as resource availability, competition, and cellular interactions—on the fitness and distribution of karyotypes. Complementing these simulations, we are developing a bioinformatics pipeline for spatial transcriptomics data from glioblastoma clinical samples. This dataset includes samples from various regions (normal, tumor, and tumor-infiltrating) across 31 samples from 19 patients, using 10X Visium with an average of 3, 100 spots per sample. The pipeline is designed to provide detailed information about the spatial organization of karyotypic diversity, elucidating how aneuploid karyotypes are distributed and interact within the tumor microenvironment. By integrating this spatially resolved data with the ABM, we aim to generate a comprehensive framework to explore how microenvironmental pressures in glioblastoma tumors shape the evolution of karyotypic diversity and drive the emergence of therapeutic resistance in patient-specific contexts. Citation Format: Richard J. Beck, Noemi Andor. Decoding tumor evolution through fitness landscapes of aneuploid cells [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 2486.