Background:Immune checkpoint inhibitors (ICIs) targeting programmed cell death protein 1 (PD-1) have significantly improved outcomes in patients with recurrent or metastatic head and neck squamous cell carcinoma (HNSCC). However, durable clinical benefit is restricted to a subset of patients, and robust predictive biomarkers remain lacking. Beyond PD-L1 expression, increasing evidence suggests that the spatial organization and functional state of the tumor immune microenvironment (TIME) critically influence response to immunotherapy. Digital Spatial Profiling (DSP) enables high-plex, compartment-resolved analysis of immune and tumor biomarkers directly within tissue architecture, offering a powerful approach to dissect spatial immune heterogeneity.
Objectives
To characterize the spatial immune landscape of HNSCC and identify protein signatures associated with response to PD-1 blockade using a high-resolution DSP approach.
Methods
Tumor tissues were retrospectively collected from 12 patients with HNSCC treated with nivolumab at the IRCCS “G. Pascale” Cancer Institute (Naples, Italy). Patients were stratified according to RECIST criteria into responders (complete/partial response or stable disease ≥6 months) and non-responders (progressive disease or stable disease <6 months). Digital Spatial Profiling was performed using the GeoMx® NanoString® platform with a 77-protein immune-oncology panel. Regions of interest (ROIs) were manually selected by expert pathologists and segmented into four molecular compartments: tumor epithelial cells (PanCK?), leukocytes (CD45?), macrophages (CD68?), and stromal cells (SYTO13?/PanCK?). Following normalization and log transformation, univariate non-parametric analyses were conducted using the Mann–Whitney U test with false discovery rate (FDR) correction.
Results
Responders exhibited a distinct spatial immune architecture characterized by increased infiltration of functionally active CD8? T lymphocytes within tumor (PanCK?) compartments. These T cells showed co-expression of CD127 (IL-7R) and granzyme B, consistent with a preserved cytotoxic effector phenotype despite PD-1 expression (p < 0.05). In CD45? immune regions, responders demonstrated coordinated upregulation of T-cell and antigen-presenting cell markers, including CD3, CD45RO, HLA-DR, CD11c, CD14, CD68, CD163, and CD56 (FDR-adjusted q < 0.1), suggesting an integrated adaptive and innate immune response. At the tumor–stromal interface, enrichment of STING and LAG-3 was observed in responders, indicating activation of innate immune sensing pathways coupled with adaptive immune regulation. In contrast, non-responders displayed a more immunologically inert or spatially disorganized microenvironment. Notably, pronounced intra-compartment heterogeneity was detected within the macrophage population, highlighting the complexity of immune modulation in HNSCC.
Conclusions
Spatial proteomic profiling reveals that response to PD-1 blockade in HNSCC is associated with a coordinated and compartment-specific immune architecture rather than isolated biomarker expression. These findings support the relevance of spatially resolved immune biomarkers as predictive tools and provide a biological rationale for integrating DSP-based approaches into immunotherapy stratification strategies. Ongoing spatial transcriptomic analyses aim to refine gene-level signatures and further strengthen the translational potential of this approach.