Multiomic analysis of treatment-naïve NSCLC before the era of neoadjuvant immunotherapy reveals contrasting immune phenotypes in stage IIIA: node-dominant (T1N2) exhibiting hot versus tumor-dominant (T4N0) cold features.

[BACKGROUND] Non-small cell lung cancer (NSCLC) exhibits substantial heterogeneity in its tumor immune microenvironment, which critically influences therapeutic outcomes. Recently, immune checkpoint inhibitors (ICIs) have been increasingly incorporated not only in metastatic settings but also into neoadjuvant, adjuvant, and perioperative treatments. Pathologic complete remission (pCR) following neoadjuvant immunotherapy is strongly associated with reduced recurrence and favorable survival outcomes. However, reliable biomarkers for predicting preoperative immunotherapy response remain limited. The dichotomy between "hot" and "cold" tumors provides a useful framework to explain differential immunotherapy responsiveness, yet how these immune phenotypes manifest across specific tumor stages remains poorly defined.

[METHODS] We performed integrative analyses combining bulk RNA sequencing and immunohistochemistry (IHC) on tumor specimens from patients with stage T1N2 and T4N0 NSCLC. Frozen samples were used for RNA sequencing and formalin-fixed paraffin-embedded (FFPE) tissues for IHC, obtained from surgeries performed between 2009 and 2018-prior to the widespread introduction of immunotherapy. Differentially expressed genes (DEGs) were identified, and Gene Ontology (GO) enrichment analyses were conducted to characterize biological pathways. Immune cell infiltration and spatial distribution were assessed by IHC staining for CD3, CD4, CD8, CD56, FOXP3, CD68, and CD163 markers. Statistical analyses were performed using the Mann-Whitney U test.

[RESULTS] Transcriptomic and histologic analyses revealed distinct molecular and immune profiles between T1N2 and T4N0 tumors. T1N2 tumors were enriched for immune activation pathways and displayed dense, evenly distributed infiltration of CD3, CD4, and CD8 T cells. In contrast, T4N0 tumors showed upregulation of extracellular matrix organization and adhesion-related pathways, accompanied by reduced intratumoral CD4 and CD8 T-cell density, accumulation of FOXP3 regulatory T cells, and increased CD163 M2 macrophages, consistent with an immune-excluded phenotype.

[CONCLUSION] Our integrative multiomic analysis delineates node-dominant (T1N2) NSCLC as "hot" tumors with active immune engagement, whereas tumor-dominant (T4N0) NSCLC exhibit "cold" features characterized by stromal remodeling, immune exclusion, and immunosuppression. These findings suggest that the balance between tumor invasiveness and nodal spread may shape immune contexture and potentially modulate responsiveness to neoadjuvant immunotherapy, highlighting the clinical value of immune landscape profiling in personalized NSCLC treatment.