.png){kind=icon}
.png){kind=icon}
Return
Construction of a cross-level atlas of the myeloid response-to-resistance continuum in non-small cell lung cancer immunotherapy and its associations with T-cell dysfunction and spatial localization
.png){kind=icon}
OA
AI
Y
Z
Y
X
A
P
DOI：10.1007/s12672-026-05505-z.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Immunotherapy efficacy in non-small cell lung cancer (NSCLC) is heterogeneous, and myeloid contributions to response and resistance are incompletely organized across single-cell, bulk, and spatial layers. We integrated public NSCLC immunotherapy datasets to define marker-based myeloid states, construct response- and resistance-associated programs, and test their transferability. Program genes were selected from cluster markers using adjusted P < 1e-10, average log2 fold change > 1, non-coding/ribosomal gene filtering, and the top 15 genes per source cluster. Scores were calculated as unweighted mean z scores, and the balance index was defined as the resistance score minus the response score. Slingshot pseudotime, diffusion map embedding, and k-nearest-neighbor state-transition analysis provided quantitative support for a response-to-resistance organization. Slingshot pseudotime correlated with the resistance program (rho = 0.649), balance index (rho = 0.639), inflammatory-neutrophil module (rho = 0.638), and inversely with antigen presentation (rho = −0.733; all BH-adjusted P < 0.001). Diffusion component 1 showed concordant correlations with the resistance program (rho = 0.782) and balance index (rho = 0.755). Bulk validation was heterogeneous and should not be interpreted as an independent clinical predictor. In the external single-cell cohort GSE243013, the resistance program and balance index were higher in non-major responders (FDR = 0.0066 and 0.0067; directional AUC = 0.617 and 0.609), whereas the response program showed the expected direction but was not significant. Myeloid program balance was associated with exhausted-like and dysfunctional T-cell states, and spatial analysis showed tumor-region enrichment of the resistance program with lower cytolytic context. The data support a transferable, association-based myeloid response-to-resistance axis in NSCLC immunotherapy. The framework is complementary to established inflamed biomarkers and requires prospective and functional validation.
Keywords:
Non-small cell lung cancer
Immunotherapy
Myeloid cells
Myeloid response-to-resistance continuum
Immune microenvironment
Single-cell transcriptomics
Spatial transcriptomics
T-cell dysfunction
Journal
IF:
2.9
Papers:
3.5K
Citations:
1.6K
Organization
Cited Papers