Spatially-resolved pharmaco-ecology of immunotherapy resistance in lung adenocarcinoma: Single-cell and spatial transcriptomics for mechanism-matched combination therapies.

Abstract

Immune checkpoint inhibitors have reshaped the treatment landscape of lung adenocarcinoma, but primary, adaptive and acquired resistance are still the main clinical obstacles. This review integrates LUAD-specific evidence with broader NSCLC single-cell and spatial-omics studies, emphasizing lung adenocarcinoma where subtype-specific data are available. We discuss how therapeutic stress can select or induce malignant epithelial states that manifest as impaired antigen presentation, altered interferon signaling, partial epithelial-mesenchymal transition, stress tolerance, and hypoxia-related metabolic adaptations. These programs interact with dendritic cells, exhausted and precursor T cells, regulatory T cells, B cells, tertiary lymphoid structures, myeloid cells, cancer-associated fibroblasts, extracellular matrix, and blood vessels to generate a unique resistance-associated spatial immune niche. Special emphasis is given to the niche with adaptive suppression, immunologically excluded interstitial boundaries, myeloid-rich hypoxic-metabolic suppressive niches, antigen-presentation-low residual niches, and tertiary lymphoid structure (TLS)-associated immuno-tissue niche. We further highlight the translational potential of integrating single-cell and spatial features for biomarker discovery, prognostic assessment, interpretation of residual disease, and mechanism-matched combination therapies. Finally, we outline key challenges for clinical implementation, including sampling bias, platform heterogeneity, distinguishing association from causation, and the need for FFPE-compatible testing and prospective validation. Spatially informed single-cell analysis may enable lung adenocarcinoma to shift from descriptive immunophenotyping to ecological diagnosis and precision immune regulation.