Metabolic-immune rewiring and spatial dynamics in therapy-resistant cancers: an AI-integrated framework for next-generation immunotherapy

Immunotherapy resistance remains the central challenge in modern oncology. While individual resistance mechanisms — metabolic exhaustion, immune exclusion, and antigen escape — have been extensively studied in isolation, the field lacks an integrated framework that connects these dimensions. This Research Topic will establish precisely such a framework by targeting the intersection of three rapidly maturing fields: tumor metabolic reprogramming, spatial immune architecture, and AI-assisted therapeutic design.Recent breakthroughs now make this integration feasible. Spatial transcriptomics and proteomics (e.g., Visium HD, Xenium, CODEX) have revealed that immune cell positioning within the tumor microenvironment — not just abundance — dictates therapeutic outcome. Concurrently, metabolomic profiling has demonstrated that metabolic competition (lactate shuttling, lipid peroxidation, tryptophan catabolism via IDO/TDO) actively drives T-cell exclusion and exhaustion. Meanwhile, AI and deep learning are transforming drug discovery, enabling the prediction of neoantigen immunogenicity, the rational design of optimized CAR constructs, and the modeling of metabolic-immune networks. Yet, no existing Research Topic connects these pillars as a unified framework.We invite Original Research, Reviews, Methods, and Brief Reports addressing:1. Metabolic Reprogramming Driving Immune Evasion• Nutrient competition (glucose, glutamine, lipids) and T-cell/NK-cell dysfunction• Lactate and proton dynamics in immune cell trafficking and polarization• Kynurenine/AhR pathway signaling in resistance• Ferroptosis and immunogenic cell death• CAF and adipocyte metabolic crosstalk with immune cells2. Spatial Immune Architecture of Resistance• Spatial omics (transcriptomics/proteomics) mapping immune desert and exclusion phenotypes• Tertiary lymphoid structures in sustaining immunity under checkpoint blockade• Metabolic enzyme distribution relative to immune cell positioning• Heterogeneity of antigen presentation machinery across tumor subregions3. AI and Computational Design for Therapeutic Discovery• Machine learning models for immunotherapy response prediction from multi-omics data• AI-guided neoantigen vaccine and TCR therapy design• Deep learning for protein engineering (CAR optimization, bispecific antibodies)• Network modeling of metabolic-immune interactions for combinatorial target discovery4. Metabolism-Engineered Cellular Therapies• Metabolism-enhanced CAR-T, CAR-NK, and CAR-Macrophages (engineered to resist lactate, maintain oxidative phosphorylation)• Logic-gated synthetic biology circuits responsive to TME signals• CRISPR base editing and prime editing for allogeneic, off-the-shelf cell therapies• Armored CARs with cytokine-secreting or stroma-targeting payloads Topic editors: Pengpeng Liu, Li Yu Source: https://www.frontiersin.org/research-topics/82356/metabolic-immune-rewiring-and-spatial-dynamics-in-therapy-resistant-cancers-an-ai-integrated-framework-for-next-generation-immunotherapy