Abstract
Abstract PD-1/PD-L1 blockade can produce durable tumor control, yet primary, adaptive, and acquired resistance remain common. Existing accounts often catalogue resistance by cellular compartment, obscuring the coordinated nature of tumor adaptation. Here, we introduce immune-pressure redistribution as a treatment-oriented framework that complements cancer immunoediting by asking where therapeutic immune pressure is diverted after checkpoint release. Resistance is organized into three coupled routes: transfer into tumor-intrinsic escape through antigen-presentation loss, interferon-response defects, oncogenic rewiring, and lineage plasticity; weakening through defective priming, terminal T-cell differentiation, compensatory checkpoints, metabolic constraint, and chronic cytokine signaling; and unloading into stromal, vascular, myeloid, regulatory, microbial, and systemic host compartments. We integrate clinically validated mechanisms with emerging evidence, including the temporal duality of interferon–JAK signaling, the role of tumor-draining lymph nodes in sustaining progenitor-exhausted T cells, and the limited translation of TIGIT, IDO1, TGF-β, and CSF-1R targeting. We further propose a biomarker-guided strategy that combines tumor visibility, immune-cell state, spatial architecture, systemic inflammation, and early treatment dynamics to identify the dominant resistance topology. This framework supports topology-matched combinations and adaptive sequencing rather than uniform escalation, with the aim of restoring productive immune pressure while limiting compensatory escape and toxicity. Graphical abstract