Abstract Background: Glioblastoma is largely resistant to immune checkpoint blockade (ICB), unlike melanoma where patients routinely derive benefit. To investigate the mechanistic basis of this divergence, we applied a predator-prey mathematical framework, treating immune cells as predators and cancer cells as prey, to model co-evolution of tumor and immune populations and to identify key parameters that restrict effective immune control in glioblastoma. Methods: We constructed coupled ordinary differential equations (ODEs) representing tumor cells, pro-tumor immunosuppressive cells (tumor-associated macrophages and myeloid-derived suppressor cells), and CD4+ and CD8+ tumor-infiltrating lymphocytes (TILs). The TIL compartment was modeled along an exhaustion continuum (activated → progenitor-exhausted → terminally exhausted), reflecting functional decline observed in glioblastoma. Recruitment, proliferation, and infiltration of effector T cells were governed by neoantigen-scaled influx terms, while tumor killing and suppressive feedback were encoded through nonlinear interaction terms. Parameter sweeps and Sobol global sensitivity analyses identified rate constants that most strongly influence tumor control versus immune escape. The model will be calibrated with melanoma immune-phenotyping flow cytometry data to benchmark the “cold” glioblastoma ecosystem against melanoma’s “hot,” inflammation-high state. Results: Simulations recapitulated clinical behaviors, including unchecked tumor growth under dominant immunosuppression and transient regressions followed by relapse driven by T-cell exhaustion. Sensitivity analyses identified exhaustion-transition rates and T-cell recruitment efficiency as the key bifurcation parameters separating durable immune control from escape. Comparison showed that glioblastoma occupies a constrained, T-cell-poor, low-inflammation parameter space that severely limits T-cell recruitment and blunts immunotherapy responsiveness, whereas melanoma resides in a more permissive, inflammation-driven regime. The framework is being extended to simulate ICB and exhaustion-reversal strategies. Visualization outputs will include therapeutic response trajectories, parameter-to-phenotype maps, phase-space comparisons, parameter-regime heatmaps, and time-course simulations of tumor, suppressive myeloid cells, and CD8+ TIL subsets across exhaustion states. Conclusions: This model integrates ecological dynamics with translational immunology to define quantitative constraints that underlie the failure of glioblastoma to respond to ICB. Mapping computational parameters to patient-derived immune features provides a predictive tool for rational combination therapy design and biomarker discovery aimed at overcoming immune escape. Citation Format: Jodie Jepson, Elizabeth Owens, Elise Nackley, Helen Rizos, Georgina V. Long, Mustafa Khasraw. Contrasting immune-tumor (predator-prey) dynamics in glioblastoma and melanoma to explain divergent responses to checkpoint blockade 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 6841.
Jepson et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: